A question about woofers
Jonathan Fischer
As I've been listening to speakers over the last few months,
it occurred to me to ask a silly question, but one which has been
nagging me lately. Don't misconstrue that I'm a Cerwin Vega Bass-Head
or anything of the sort...
Just why have large woofers fallen so out of fashion in
high-end speakers [or have they ever been in fashion in the high-end]?
The biggest thing I've been seeing in the speakers I've been looking
at is around 8", with a 6 - 6 1/2" woofer being very common. There's
the odd 10" or 12" woofer out there, but these are an exception, at
least in my experience. Even when there is a third, midrange, driver,
the woofer is still usually no larger than 8". My 10-year-old
speakers have a 12" woofer, and that was very common back then.
What are the technical disadvantages of a large woofer, and/or
what are the technical advantages of a small woofer?
Also, while I'm at it, what exactly is the transmission line
speaker? Is it just a fancy porting method?
--
Jonathan A. Fischer SCO Canada, Inc.
jfis...@scocan.sco.COM Toronto, Ontario, Canada
Usenet's first law of flamodynamics:
For every opinion, there is an equal and opposite counter-opinion.
Dave Neff
This reminds me of a similar question about woofers (I don't mean
to start drift, but perhaps the answers to the questions are related).
I noticed that the new Advent Heritage speakers have two 8" woofers in
each speaker. Not only are the woofers small, but it has 2 of them
and they appear to be identical. What would be the advantage of 2
(apparently identical) woofers in a single speaker?
Dave Neff
ne...@hpvcfs1.HP.COM
David Chin
are that small woofers tend to be more accurate. For the same material
and hence stiffness, a smaller cone will be less prone to breakup (i.e.
different parts of the cone move slightly differently, sometime out of
phase even). Cone breakup means distortion. Also, smaller cones have less
mass and hence better transient response. I doubt whether this is very
audible though. What may be more audible is that the the transient response
of a small woofer matches better the fast transient of midrange drivers and
the much faster transient response of tweeters. The fact that some
speakers align their drivers so that the woofer is forward of the midrange
which in turn is forward of the tweeter may ameliorate this transient
response matching problem. It also appears to be a big problem in
matching electrostatic speakers (extremely fast) to moving coil woofers.
This may also explain why many two-way speakers have better sound stages
than three way speakers (transient information is more important for
locating sounds, hence a mismatch in transient response between woofer
and midrange and tweeter muddies the soundstage).
Finally, the reason one uses two woofers is to move more air (more air
means louder). For the same excursion limits (maximum peak to peak travel
of the woofer), two 8" woofers will be as loud as one 11.3" woofer (at
least theoretically based on 2*pi*r*r). Why not just build woofers with
longer throws (another term for excursion limits)? That has to do with
the design of the surrounds (the rubbery part that holds the woofer in
place). It is hard to design surrounds that will hold a cone and coil
in place correctly with more than about 1" of excursion.
Disclaimer: I am not an audio engineer, so some of the above may be
incorrect, in which case I would appreciate corrections.
Dave
William Spencer
> Just why have large woofers fallen so out of fashion in
> high-end speakers [or have they ever been in fashion in the high-end]?
Actually they're not out of fashion for those into subs. The Wilson
WHOW, for example, uses an 18". But you're not seeing a lot of large subs
in most shops.
But when you're talking a regular woofer that has to handle higher frequencies
it must be small.
> Even when there is a third, midrange, driver,
> the woofer is still usually no larger than 8".
A lot of high end 3-ways have a fairly high crossover between woofer and mids.
That's so you can use the specialized drivers such as dome midranges and
certain tweeters that are fairly limited in range to get that ultra-detailed
sound. In my opinion, the load should distributed evenly between drivers.
For a three way system with 20-20K response that would be frequencies of
200 and 2K (3.33 octaves/driver). For two way the xover would be
666 Hz (ooh, scary) (5 octaves/driver), not
possible with common drivers. The 200 is sort of bad because it's
a frequency rather sensitive to phase changes. If anything, the mid should
have the widest range because the extremes are more difficult to reproduce.
> My 10-year-old
> speakers have a 12" woofer, and that was very common back then.
And not uncommon in subs now.
> What are the technical disadvantages of a large woofer, and/or
> what are the technical advantages of a small woofer?
Large woofers can lack transients and mids. But, really, anything
you can do with a large woofer you can do with multiple small ones,
at least in theory. The suspension and excursion may be lacking, but
not as much with certain small expensive drivers. You can also do
tricks like push-pull with multiples. But the biggest change is
a movement towards less resonant tunings and smaller boxes. You must
have a big box for
big woofers to sound right. And big boxes can cause trouble.
Bill Spencer
Patrick W Lauson
>The theories that I hear (has anyone on the net validated them empirically?)
>are that small woofers tend to be more accurate. For the same material
>and hence stiffness, a smaller cone will be less prone to breakup (i.e.
>different parts of the cone move slightly differently, sometime out of
>phase even). Cone breakup means distortion. Also, smaller cones have less
This still doesn't change the fact that larger woofers can reach lower
frequencies... Also, cones, can be dampened to reduce breakup modes..
I personally prefer 12's or 15's for the lower bass, but I guess
that it's all up to the individual...
BTW... Have you seen a speaker in breakup? Pretty neat..
>
> Dave
--
Later,
Pat Lawson
lau...@ecn.purdue.edu
Francis Vaughan
(David Chin) writes:
|> Disclaimer: I am not an audio engineer, so some of the above may be
|> incorrect, in which case I would appreciate corrections.
Nor am I, but I think a couple a statements are a bit misleading. Non the
less I perfectly happy to have my corrections corrected.
|> The theories that I hear (has anyone on the net validated them empirically?)
|> are that small woofers tend to be more accurate. For the same material
|> and hence stiffness, a smaller cone will be less prone to breakup (i.e.
|> different parts of the cone move slightly differently, sometime out of
|> phase even). Cone breakup means distortion. Also, smaller cones have less
|> mass and hence better transient response.
No trouble with this statement. Nicely stated.
|> I doubt whether this is very
|> audible though.
Depends upon how good a sound quality you expect. After a certain point I
feel this does start to make a difference. Of course the crossover imposes
a maximum rise time. With low order crossovers (which are favored by those
designers that worry about transient response) the drivers rise time becomes
more important.
|> What may be more audible is that the the transient response
|> of a small woofer matches better the fast transient of midrange drivers and
|> the much faster transient response of tweeters.
This is a bit non-sensical I think. If a bass drivers transient response was
the same as a tweeters then it would be a full range driver. (Diffraction
problems aside.) To say that the driver sould have the best rise time one
can, is a restatement of the previous comments.
|> The fact that some
|> speakers align their drivers so that the woofer is forward of the midrange
|> which in turn is forward of the tweeter may ameliorate this transient
|> response matching problem.
I suspect you are confusing issues in phase response with those of rise time.
Aligning drivers at different levels on the baffle is an attempt to align
the acoustic centers of the drivers (a quite usless rule sometimes used, is
to align the dust caps of the drivers in the belief that they were the acoustic
centers.) This results in a correct phase alignment of the drivers in the
cross over regions and prevents the "lobing" effect if used with appropriate
crossover sections (Riley Linkwitz is the best known). The mechanical design
of the driver can place the acoustic center anywhere within limits. Dynaudio
drivers for instance, are specificly engineered to align the acoustic centers
when all the drivers are mounted on a flat baffle.
|> It also appears to be a big problem in
|> matching electrostatic speakers (extremely fast) to moving coil woofers.
|> This may also explain why many two-way speakers have better sound stages
|> than three way speakers (transient information is more important for
|> locating sounds, hence a mismatch in transient response between woofer
|> and midrange and tweeter muddies the soundstage).
Probably much more to do with the lack of a defined acoustic center of
the speaker
with drivers spaced futher apart, and problems with a bigger box, and often
poorer quality drivers compared to a similar two way. Very big speakers
can have
superb imaging, but is is a lot harder to achieve than with small ones. The
symetric placement of mids and bass drivers above and below the tweeter is
probably the best solution (and probably the most expensive). In principle
with appropriate crossovers and alighment of drivers you can make it seem to
the ear that is is the center of the world acousticly. This is all acheived
by careful design, watching the phase angles.
With any luck, there is scope for more discussion in the above ramblings.
Francis Vaughan
Len Moskowitz
> What would be the advantage of 2 (apparently identical) woofers in
> a single speaker?
They move more air so you get higher sound pressure levels (i.e., it's
louder).
Len Moskowitz
> This still doesn't change the fact that larger woofers can reach
> lower frequencies...
This is a common misconception. You can make a 4" driver put out 20
Hertz tones; they just won't be very loud. Low frequency response
(not sound pressure levels) is solely a result of proper design a
driver's key parameters, and proper matching of that driver with its
cabinet.
Small drivers can go down to very low fequencies but need huge
excursions to put out reasonable sound pressure levels. I remember a
4" driver that had almost an inch of excursion. It sounded pretty
amazing at low SPLs if you neglected the doppler distortion. In most
cases, when you need really low bass, a better tradeoff is to use a
larger driver.
Len Moskowitz
The Friend
Additionally you can wire them in parallel, getting half the normal
ohm rating, yet doubling power handing. Given your amp can handle the low
impedence (4 ohms typical), you can both get better SPLs, and at the
same time put more power through the system..
Shankar Bhattacharyya
> > This still doesn't change the fact that larger woofers can reach
> > lower frequencies...
>This is a common misconception. .................
>Small drivers can go down to very low fequencies but need huge
>excursions to put out reasonable sound pressure levels. ............
To emphasize which, consider that a closed box woofer has to be able to
displace almost 2 liters of air, peak to peak, to produce 100 dB at 30 hz
at 1 meter from the driver. That is more displacement than some automobile
engines have.
- Shankar
P E Smee
> What are the technical disadvantages of a large woofer, and/or
>what are the technical advantages of a small woofer?
A primary disadvantage of large woofers is that old physical friend,
inertia. (Or, its secret identity, momentum.) Larger woofers will go
lower, but at the cost (for a given technology) of some loss of
'sharpness' or precision. They are likely to go muddy, or to 'thump'
(which some people actually like) rather than making bass 'notes'.
There's added inertia in the larger cone, which means more in the
larger air mass to be displaced, and yet again more in the larger coil
needed to push these two around.
You can compensate for this by use of super-efficient voice coils and
lots of input power, but then you're talking about lotsamoney (for the
coils) and something that a lot of amps can't handle (because of the
power requirement). Modern designs with smaller woofers give a better
compromise between good sound reproduction and physical (and marketing)
reality.
--
Paul Smee, Computing Service, University of Bristol, Bristol BS8 1UD, UK
P.S...@bristol.ac.uk - ..!uunet!ukc!bsmail!p.smee - Tel +44 272 303132
david.r.moran
>In article <1991Feb20.1...@sco.COM> jfis...@sco.COM (Jonathan Fischer) writes:
>> What are the technical disadvantages of a large woofer, and/or
>>what are the technical advantages of a small woofer?
>
>A primary disadvantage of large woofers is that old physical friend,
>inertia. (Or, its secret identity, momentum.) Larger woofers will go
>lower, but at the cost (for a given technology) of some loss of
>'sharpness' or precision. They are likely to go muddy, or to 'thump'
>(which some people actually like) rather than making bass 'notes'.
>There's added inertia in the larger cone, which means more in the
>larger air mass to be displaced, and yet again more in the larger coil
>needed to push these two around.
>
>You can compensate for this by use of super-efficient voice coils and
>lots of input power, but then you're talking about lotsamoney (for the
>coils) and something that a lot of amps can't handle (because of the
>power requirement). Modern designs with smaller woofers give a better
>compromise between good sound reproduction and physical (and marketing)
>reality.
>
Shoot, I was hoping it would not get this far. Mr Smee is usually
full of wisdom and an interesting viewpoint from across the pond, but,
alas, notions of "fast" response
to transients, whatever they are, are bogus. You don't want a woofer to
be fast, or slow, you want it to respond to the input. (For good "transient"
response in the bass, actually, you need good tweeters, because the leading
edge of the slams and cracks we all love in deep bass are high frequencies...
but that is another post for another day.)
There is no advantage small woofers have over big ones except that their
dispersion is less beamy and roughened as you go higher up toward the
crossover to the midrange compared with big woofers. Big woofers can
get badly honked and chesty as they are driven higher and higher up.
So can 6"ers if you drive them up too high, a common occurrence.
This dispersion problem -- radiation pattern or coverage or directional
responses
are other terms -- is a serious matter and in the horizontal plane is
the chief cause of differences among speakers. Off-axis "scallops" or
"seams/stitches" at the crossover, where one driver flares in with
wide coverage as the
other (lower one) struggles upward and gets beamy, are best avoided but
are common and many designers pay little attention to them, if any.
Truly uneven imaging is the other chief result of this, too, as well
as honked sound.
Boom and slow, muddy, thuddy response, so called, and thud and lack of
speed (!) are in general functions of box volume loading, alignment,
room loading (by the three nearest boundaries, etc. etc.).
This is another tweak myth that has raped physical and scientific reality,
I am afraid.
Keith Harner
effectively is the ability to move lare amounts of air at the
desired frequency. I personally favor larger low frequency drivers
and my system uses 12 inch three way satellites with a pair of
15 inch subs.
I have heard speakers that use multiple drivers of a smaller size
quite effectively. (Sorry Boise, I'm talking no eq's here.)
One in particular that really impressed me was a design marketed under
the name "shot glass" it used (as I vaguely recall) five drivers
of about 6 inch diameters mounted in a vertical line source
configuration. Relatively inefficent but very good sounding.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
|Keith Harner |har...@civax.DAB.GE.COM |
|General Electric Simulation and Control Systems|har...@sunny.DAB.GE.COM |
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Norm Strong
You probably answered your own question. There are 2 of them because they
are so small that it takes 2. Why don't they use 1 larger one, you may
ask? There are many possibilities: 1. A larger woofer won't work to as
high a xover frequency; that might have necessitated a 3 way system. 2.
They got a tremendous buy on the woofers, since they already use them in
other speakers. 3. They wanted to keep the enclosure narrow.
In other words, who knows?
--
Norm Strong (str...@tc.fluke.com)
2528 31st S. Seattle WA 98144 USA
J...@slacvm.slac.stanford.edu
In article <1991Feb25.1...@tc.fluke.COM>, str...@tc.fluke.COM (Norm
Strong) says:
>
>In article <223...@hp-vcd.HP.COM> ne...@hp-vcd.HP.COM (Dave Neff) writes:
>}Regarding small woofers:
>}each speaker. Not only are the woofers small, but it has 2 of them
>}and they appear to be identical. What would be the advantage of 2
>}(apparently identical) woofers in a single speaker?
>
>You probably answered your own question. There are 2 of them because they
>are so small that it takes 2. Why don't they use 1 larger one, you may
>ask? There are many possibilities: 1. A larger woofer won't work to as
>high a xover frequency; that might have necessitated a 3 way system. 2.
>They got a tremendous buy on the woofers, since they already use them in
>other speakers. 3. They wanted to keep the enclosure narrow.
>
>In other words, who knows?
I know nothing about speaker designs but...
could it be that the smaller units are quicker due to size, yet together,
move the same amount of air as a large woofer? Would this not give you
tighter, less flabby bass then perhaps a 12" woofer? Seems right to me...
John
Cal Demaine
>
>>}each speaker. Not only are the woofers small, but it has 2 of them
>>}and they appear to be identical. What would be the advantage of 2
>>}(apparently identical) woofers in a single speaker?
>>
If it was up to me (and fortunately, it is not) I would choose a two woofer
design for a number of reasons.
1. You can cross them over at different frequencies
2. You can wire them in series/parallel to match impedences
3. You can put them in seperate chambers, thereby tuning them differently
4. You can use one as a passive radiator (kind of a dumb idea)
The important points, I believe, are 1 and 3. Most two-woofer systems I've
seen (broad generalization) do employ the ideas behind 1 or 3 or both.
- Cal
Lon Stowell
>In article <1991Feb20.1...@sco.COM> jfis...@sco.COM (Jonathan Fischer) writes:
>> What are the technical disadvantages of a large woofer, and/or
>>what are the technical advantages of a small woofer?
>
>A primary disadvantage of large woofers is that old physical friend,
>inertia. (Or, its secret identity, momentum.) Larger woofers will go
>lower, but at the cost (for a given technology) of some loss of
>'sharpness' or precision. They are likely to go muddy, or to 'thump'
>(which some people actually like) rather than making bass 'notes'.
>There's added inertia in the larger cone, which means more in the
>larger air mass to be displaced, and yet again more in the larger coil
>needed to push these two around.
>
as a mid-bass or midrange transducer. If you ever read the
set-up manual for a good SUBwoofer, it explicitly notes that
the subwoofer...operating w/o main speakers....should have
ABSOLUTELY no musical tone whatever...you should get a soft
rumbling growl with ABSOLUTELY no attack/decay transients.
These attack/decay transients...even for a pulsed 15 Hz
signal....are routed by a PROPERLY designed xover as the
higher frequency components that they actually ARE over to
the midrange and even tweeter units.
Large woofers DO have a lot of inertia...and can be
problematic for poorly designed amplifiers. The stored
energy of the cone in motion must be dissipated into the
internal impedance of the amplifier...then it must be moved
in the opposite direction.
When you get in trouble is when you forget this (and the
beaming characteristics of large cones relative to ANY
frequency) and attempt to use a large transducer above its
optimal frequencies.
>You can compensate for this by use of super-efficient voice coils and
>lots of input power, but then you're talking about lotsamoney (for the
>coils) and something that a lot of amps can't handle (because of the
>power requirement). Modern designs with smaller woofers give a better
>compromise between good sound reproduction and physical (and marketing)
>reality.
>
>--
It may likely BE economics....modern designs which attempt
to get more bandwidth out of each driver and/or keep the
crossovers out of the critical midrange tend to go with
smaller woofers for the 2 way systems now in style. Even
with a 3-way you can likely get more bang (boom?) for the
buck with a good, long excursion, smaller woofer in a
properly designed cabinet.
In a good 4-way or subwoofered system, the large woofers are
really more efficient...so pay your $$$ and let the plaster
loosen.
It may also be that modern manufacturing can produce smaller
drivers with excursions able to produce reasonable bass
SPL...this wasn't true in the recent past.
Another factor may be construction quality of housing.
Subwoofers or the old 36" Electro-Voice woofers do NOT make
good neighbors in the thinwall housing construction
techniques of these days....
Lon Stowell
>
> > This still doesn't change the fact that larger woofers can reach
> > lower frequencies...
>
>Hertz tones; they just won't be very loud. Low frequency response
>(not sound pressure levels) is solely a result of proper design a
>driver's key parameters, and proper matching of that driver with its
>cabinet.
>
>4" driver that had almost an inch of excursion. It sounded pretty
>amazing at low SPLs if you neglected the doppler distortion. In most
>cases, when you need really low bass, a better tradeoff is to use a
>larger driver.
Depends on what you want. Using about 6 of the old KLH drivers
gave pretty good bass....just don't use the same physical driver for
midrange or you will get IM distortion....as you noted.
Of course the entire issue of "FAST" woofers is pretty silly as
long as the speaker is really a woofer and not also serving as
the lower midrange driver.
Shankar Bhattacharyya
>If it was up to me (and fortunately, it is not) I would choose a two woofer
>design for a number of reasons.
> 1. You can cross them over at different frequencies
> 3. You can put them in seperate chambers, thereby tuning them differently
>The important points, I believe, are 1 and 3. Most two-woofer systems I've
>seen (broad generalization) do employ the ideas behind 1 or 3 or both.
How about an example or two of these? I have not looked at the market for a
long time, and am interested in what manufacturers are doing.
Builders tend to use dual-woofer designs to get greater output, or,
occasionally, for the somewhat smaller net volume and lower distortion you
can get from isobaric designs. I have seen some interest in dual chamber
designs, but not that much. Among builders, that is.
Thanks for any information.
- Shankar
P E Smee
On some such boxes (don't know about the Advents) only one of the
apparent woofers is driven, and the second is a dummy cone which
(apparently) is supposed to resonate in response to the air which the
driven woofer is pushing around inside the box. (And I still don't see
why they work. Feels to me like in that case the resonator should be
pretty much guaranteed to be out of phase with the driven cone, so
causing cancellation. One of the black arts, I suppose.)
Assuming both are driven, I offer a reasonable sounding explanation
which was given to me by a speaker company rep at the Bristol hi-fi
show a week ago. (I'm parrotting, so if I'm wrong I'd like to be
corrected, but I'd prefer to be corrected gently since it's not my
ideas I'm expounding. :-)
It comes down to the problem with large woofers -- physical reality.
To get good low bass, you have to move lots of air around. This can
be done with a large driver or with several smaller ones. (It can
also be simulated by creative use of ports, but I don't understand them.)
For clean sound reproduction, you want a speaker's cone to move as a
solid unit. This means it has got to be stiff, because otherwise since
it is driven from the apex you will get a wave travelling up the cone
which will muddy the sound. Allegedly, the stiffening required makes
a single large driver disproportionately heavy as compared with using
several smaller drivers, so that you can do better (efficiency wise)
with the smaller drivers. And, you win in that two 8" drivers give
you about 100 sq inches of effective surface area, which you'd actually
need a 12" driver to equal.
On the other hand, you obviously can't take this reasoning to
extremes. A huge array of lots of tiny speakers would begin to suffer
from various forms of acoustic interference between the multiple
separate sound sources.
As I said, if this is really all marketing hype I'd like to be
corrected. Sounds reasonable to me, and the person explaining it
seemed reasonably knowledgeable and trustworthy, so I offer it up as a
target.
--
Phil Stone
>
> Another factor may be construction quality of housing.
> Subwoofers or the old 36" Electro-Voice woofers do NOT make
> good neighbors in the thinwall housing construction
> techniques of these days....
I am extremely tempted to add a subwoofer to my system, but
have not done so, because I live in an upstairs apartment with
very little acoustic isolation from the apartment below. Since
I basically like my downstairs neighbors, and don't want to make
them hate me, I wonder if there is a way to minimize the negative
impact a subwoofer would have on our co-existance?
I don't play music at really loud volumes generally, but I know
how well bass carries through structures. It would seem to be
the worst set up to have the subwoofer on the floor (their ceiling!).
Would some sort of acoustic decoupling help? Can anyone imagine
how this might be done, i.e. some kind of mount for the subwoofer
that incorporated rubber isolation of some sort?
Phil Stone (ph...@eos.arc.nasa.gov | ...ames!eos!phil)
David P. Mottaz
Keith Harner
>A huge array of lots of tiny speakers would begin to suffer
>from various forms of acoustic interference between the multiple
>separate sound sources.
This is really the only part of this post that I don't buy.
I have read about an effect called acoustic coupling in
which near-field positioning of woofers actually augements
their low-frequency response on the order of around 3 to 6 db.
Anybody ever read this ?????
Biswa Ranjan Ghosh
>In article <1991Feb26.1...@gdr.bath.ac.uk> P.S...@bristol.ac.uk (Paul Smee) writes:
>>A huge array of lots of tiny speakers would begin to suffer
>>from various forms of acoustic interference between the multiple
>>separate sound sources.
>
>This is really the only part of this post that I don't buy.
>I have read about an effect called acoustic coupling in
>which near-field positioning of woofers actually augements
>their low-frequency response on the order of around 3 to 6 db.
>
Believe it. Multiple sources in which the separation between sources is
comparable to the wavelength of the sound being reproduced will suffer
from all sorts of lobing problems due to interference. Think of it this
way. Designers go to great lengths to control interference patterns in
even simple two-way systems, which occur in the crossover region where
both drivers are operating simultaneously. Now if two similar drivers
operate simultaneously over a wide frequency range, say they are arranged
vertically, then you will get lobing in the vertical direction, and the
measured (and heard) spl will vary as you move your measuring equipment
vertically. As the frequency gets higher, the lobes become narrower and
more numerous.
This is really very much analogous to antenna radiation pattern theory.
The mathematics is beyond me (ask an E&M guru). But the theory also explains
the radiation pattern of line sources, and the difference in power drop-
off of point sources vs. line sources. (Many planar speakers use line
sources for the tweeter. Also, some of them suffer lobing effects, often
referred to as a "venetian blind" effect.)
This doesn't mean much if you are using two 12" woofers, one atop the
other, since they probably won't operate above 100 or 200 Hz., and at
those wavelengths, their separation is small enough to consider them as
a single virtual point source. But try using numerous 4" or 6" woofers
upto a Khz, and you will hear anomalies.
I'm not aware of the acoustic coupling effect you mention. But of course,
if you have two drivers in parallel, you see a greater output level for the
same input voltage since you have twice the radiating area of a single
driver. You also get greater maximum spl level, since at the maximum excursion
limits, you have twice as much volume displacement over a single driver.
(This last factor is a major consideration for subwoofers, and why I think
a compound driver arrangement in which both drivers contribute to sound
output is much preferred to isobarik configurations.)
Biswa Ghosh
550 Cory Hall arpa: bi...@janus.berkeley.edu
U.C. Berkeley uucp: ...ucbvax!janus!biswa
Berkeley, CA 94720 tel: (415) 642-0395
Mike Hejl
>
> a single virtual point source. But try using numerous 4" or 6" woofers
> upto a Khz, and you will hear anomalies.
>
--
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+
| Mike Hejl he...@necssd.NEC.COM, {necntc, cs.utexas.edu}!necssd!hejl |
| #include <stupid.disclaimer.h> |
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+
Terry Ingoldsby
> are that small woofers tend to be more accurate. For the same material
I suspect this is true under certain conditions. Surely the physical
distortion of the cones is less, all other factors equal. The mass
of the cone should be less, so that should make it easier to control.
>
> Finally, the reason one uses two woofers is to move more air (more air
> means louder). For the same excursion limits (maximum peak to peak travel
> of the woofer), two 8" woofers will be as loud as one 11.3" woofer (at
> least theoretically based on 2*pi*r*r). Why not just build woofers with
> longer throws (another term for excursion limits)? That has to do with
> the design of the surrounds (the rubbery part that holds the woofer in
> place). It is hard to design surrounds that will hold a cone and coil
> in place correctly with more than about 1" of excursion.
Traditionally, those who made speakers with a small cone compensated for
the lack of air volume by resorting to long throws. This, unfortunately
requires not only a long throw, but also high cone velocities. This not
only makes the cone hard to control, but introduces doppler distortion.
My suspicion is that by using multiple small cones it is not necessary
to resort to long throws (high velocities) to achieve the same air
movement. Control is improved and doppler distortion avoided. The
only thing that is not clear is that the absolute low end is as low.
The resonant frequency is usually higher than a large woofer (output
drops off sharply below resonance). You also get high power handling
capacity with low powered woofers.
These things seem to come and go. An old Philips book on building
loudspeakers from about 15 years ago was quite keen on them.
Anyway, these are my guesses.
--
Terry Ingoldsby ingoldsb%cty...@cpsc.ucalgary.ca
Land Information Services or
The City of Calgary ...{alberta,ubc-cs,utai}!calgary!ctycal!ingoldsb
Jerry E. Bulyk
I have a spare Klipsh 15" woofer. I would like to put it to some use. Any ideas?
some background:
I have a pair of Klipshorns that I use as subwoofers. - obvious?
more background:
I also have a pair of Accoustat speakers. I love their sound but they are
definitely lacking in the area of dynamics. ..so, I decided to add a subwoofer.
Before I started visiting stores, I found a pair of 1973 vintage Kornerhorns
locally in very good condition for only $750. With Kornerhorns I have a dual
purpose solution: a subwoofer and a general purpose _loud_speaker. I added an
active crossover and extra amp, and now have all the advantages of subwoofers.
There may very well be real subwoofers that go lower, but, there is something
very tactile to a horn-loaded woofer. (efficient too)
Last summer, I managed to push the 'horns (in standalone full range mode)
too far and blew one of the woofers. The folks at Klipsh got me a pair of their
latest rev woofers in a couple of days, leaving me with an extra 15" driver.
It really seems like a waste not to use the darn thing. I'm starting to think
in terms of a sub-sub-woofer now. (the x-over can be used as a mono three-way
x-over) hmm... maybe if I got a second x-over ... and another amp ...
- jerry
INTERNET: jbu...@telesci.uucp
UUCP: ...!princeton!telesci!jbulyk
P E Smee
>In article <1991Feb22.1...@gdr.bath.ac.uk> P.S...@bristol.ac.uk (Paul Smee) writes:
>>A primary disadvantage of large woofers is that old physical friend,
>>inertia. (Or, its secret identity, momentum.) Larger woofers will go
>>lower, but at the cost (for a given technology) of some loss of
>>'sharpness' or precision. They are likely to go muddy, or to 'thump'
>>(which some people actually like) rather than making bass 'notes'.
>>There's added inertia in the larger cone, which means more in the
>>larger air mass to be displaced, and yet again more in the larger coil
>>needed to push these two around.
>>
>>You can compensate for this by use of super-efficient voice coils and
>>lots of input power, but then you're talking about lotsamoney (for the
>>coils) and something that a lot of amps can't handle (because of the
>>power requirement). Modern designs with smaller woofers give a better
>>compromise between good sound reproduction and physical (and marketing)
>>reality.
>
>to transients, whatever they are, are bogus. You don't want a woofer to
>be fast, or slow, you want it to respond to the input. (For good "transient"
>response in the bass, actually, you need good tweeters, because the leading
>edge of the slams and cracks we all love in deep bass are high frequencies...
>but that is another post for another day.)
Did I say 'fast'? (I might have, but not in the bit you quoted.) No,
I know about Fourier as well. :-) What I had in mind was not that more
inertia makes speakers less able to handle transients, but that it can
make them lag in responding to the input they are supposed to be
responding to.
However, I seem to be holding a tenuous position here, so I'll withdraw
the idea. With the comment that it looks like I may have been taken in
by (believable sounding) manufacturer's hype. (Sigh, I hate it when
that happens. At least it wasn't a manufacturer I've bought anything
from.)
Jonathan Fischer
Well, it's nice to get so much response. The factors affecting
woofer cone size seem to be:
(1) Smaller cones have better transient response (speed).
(2) Smaller cones can be more rigid, thus minimizing waves
along the surface of the cone.
(3) Larger cones move more air, which is important for longer
wavelengths (lower frequencies).
(4) Two smaller cones can equal a larger cone in terms of volume of
air moved, and have the benefit of being "faster", but suffer
from some adverse side effects due to the multi-point source.
I was aware of all but (1), though it is rather obvious after some
thought. However, it took me a while to appreciate the "need for speed" in
a low-frequency driver. I mean, if something reproducing a 40Hz wave is a
little "slow", it would be off by a few degrees of phase at worst, and
phase information is entirely unimportant at that range, from what I
understand (I do like to qualify my statements, don't I?). Then I realized
that if the bass driver is too slow, it will suffer when reproducing the
impact of a bass drum (or the explosions on Telarc's "Ein Straussfest" :o) ).
This is why, I suppose, there is so much effort put into "active servo"
technology in subwoofers a la Velodyne, to keep the large suckers fast.
The speakers I am currently in love with are the Museatex "Melior
One". They are a fairly moderately-priced planar speaker with incredible
imaging and a very natural midrange and high-range (although the frequency
response curve between 100 and 1000 Hz is anything but flat). The low
range, however, drops off quickly below about 50Hz, and they're a perfect
candidate for a subwoofer. Thus my interest in woofers (I was curious as
to how easy it would be to take a 15" driver, a very low and steep (40Hz
@12dB/octave?) low-pass filter, and come up with a good subwoofer for these
speakers).
I guess the reasons I love these speakers so much are: that they
are the first planar speakers to which I've listened seriously; for planar
speakers they're not too expensive (CDN$2500); with their rosewood bases
and angular shape they're absolutely gorgeous; and they're Canadian (it's
not so much that I'm nationalistic, but domestic speakers are naturally a
good value and Canadian speakers are generally highly acclaimed). Still,
their midrange response isn't as flat as I'd like, and I'd like to compare
them to, say, the Energy 22.3 at the same price range (which does _not_
need a subwoofer anywhere near as much, and whose response curve is
far more flat).
--
Jonathan A. Fischer SCO Canada, Inc.
jfis...@scocan.sco.COM Toronto, Ontario, Canada
Usenet's first law of flamodynamics:
For every opinion, there is an equal and opposite counter-opinion.
Terry Ingoldsby
> In article <11...@uhccux.uhcc.Hawaii.Edu>, ch...@uhccux.uhcc.Hawaii.Edu (David Chin) writes:
> > The theories that I hear (has anyone on the net validated them empirically?)
> > are that small woofers tend to be more accurate. For the same material
>
> My suspicion is that by using multiple small cones it is not necessary
> to resort to long throws (high velocities) to achieve the same air
> movement. Control is improved and doppler distortion avoided. The
> only thing that is not clear is that the absolute low end is as low.
> The resonant frequency is usually higher than a large woofer (output
> drops off sharply below resonance). You also get high power handling
> capacity with low powered woofers.
Due to my haste to catch my train, I forgot to mention the most important
thing. I've noticed that current dome midrange drivers don't go as low
as midranges used to. In fact, sometimes there is very little overlap
with the response of a 12" woofer. 8" woofers usually go 1 or 2 octaves
higher than 12" drivers. This increases the overlap in the midrange/woofer
region and makes crossover smoother and easier. Crossing the midrange
higher may also raise the effective power rating of the system.
Al Peterman
speed and accuracy of large woofers. It is interesting that little
mention has been made of using motional feedback to put the speaker into
the feedback loop. This has been done by various companies for nearly
20 years, and 3 of my speaker systems use feedback. I find it gives fast,
tight bass, and lots of it from both small and larger cones.
Two of my systems use the old LWE speakers that were developed in the
early 70's. In these speakers a feedback signal is developed that is
coupled to the negative feedback circuit in the driving amplifier. What
is fun is switching off the feedback and hearing the bass quality deteriorate.
Another of my systems uses bass cones with 2 coils, and one of the coils
is devoted to correction signals derived by the crossover. This was
termed the Watkins dual drive woofers and develops a very tight, yet
powerful bass, albeit at some cost in efficiency. Luckily high current
amps are not too hard to find these days.
Using feedback is particularily suitable for woofer applications. It
is easier to deal with the relatively low frequencies with minimal phase
error. I am suprised that more companies don't use this technology
in integrated speaker systems, although there are several subwoofer
with integrated amps that still use this at the present. Of course the
IRS series V does use this in their woofer amp - but at $50,000 it's
a bit pricey.
--
Alan L. Peterman (503)-684-1984 hm
a...@qiclab.scn.rain.com
Henry Kwan
>Maybe they just wanted 10" wide cabinets. :-)
Actually, that is not as unusual as you might think. The NHT Model II's
employ two 6.5" woofer units instead of a larger driver because they wanted
in part to keep the narrow 8" (or so) wide cabinet form factor.
--
Henry Kwan | AppleLink: D0690 [] Bay BBS: (415)775-2384
FWB, Inc. | CompuServe: 71320,1034 [] FAX: (415)775-2225
2040 Polk St. Ste 215 | Internet: claris!wet!lo...@ames.arc.nasa.gov
San Francisco, CA 94109 | UUCP: {claris,hoptoad,lamc,ucsfcca}!wet!logic
Terry Ingoldsby
> On some such boxes (don't know about the Advents) only one of the
> apparent woofers is driven, and the second is a dummy cone which
> (apparently) is supposed to resonate in response to the air which the
> driven woofer is pushing around inside the box. (And I still don't see
> why they work. Feels to me like in that case the resonator should be
> pretty much guaranteed to be out of phase with the driven cone, so
> causing cancellation. One of the black arts, I suppose.)
The second cone is known as a passive radiator. This is essentially
a base reflex (a.k.a. phase inversion speaker) that (allegedly) has
a better wavefront than an ordinary port. At most frequencies (say
an octave above resonance) the radiator has little or no effect (like
a phase inversion speaker. Near resonance the speakers move in phase
and output is improved. Below resonance the drivers are in anti-phase
and output is decreased. This is a very *approximate* explanation.
For clarifications, please see the papers by Thiel which describe
this behaviour much more quantitatively.
Terry Ingoldsby
> In article <1991Feb26.1...@gdr.bath.ac.uk> P.S...@bristol.ac.uk (Paul Smee) writes:
> >A huge array of lots of tiny speakers would begin to suffer
> >from various forms of acoustic interference between the multiple
> >separate sound sources.
>
> This is really the only part of this post that I don't buy.
> I have read about an effect called acoustic coupling in
> which near-field positioning of woofers actually augements
> their low-frequency response on the order of around 3 to 6 db.
Buy it! :^) If the frequency (actually wavelength) of the sound is
small compared to the size of the radiator, then you are going to
begin to experience defraction/interference problems. As long as
the radiators are sufficiently close together to avoid these
conditions then it will not be noticable. For woofer frequencies
two 8" radiators close together are probably not a problem. If
then are stacked vertically, then you will actually have a bit
better dispersion horizontally than a single 12" woofer, and a bit
worse vertically. Dispersion is not usually a problem for woofers
so I doubt it matters at all. On the other hand, if you are using
8" woofers so that you can push the crossover point up to 2000 Hz
then you may well perceive the difference.
If this doesn't make sense it is because it is late on a Friday
afternoon.
Mithat F Konar
>
> Well, it's nice to get so much response. The factors affecting
>woofer cone size seem to be:
>
> (1) Smaller cones have better transient response (speed).
Er..no. Remember Physics 1 at college. You'll need to keep it in mind for
the following discussion.
You can generally break down the way a woofer works into two ranges: the low
frequencies where the cone is moving pistonically and the higher frequencies
where it is not. While it is true that smaller woofers extend the range
of pistonic motion (cone mass being only one variable here), it is not true
that a smaller (lower mass) cone will be "faster" in the pistonic low freq-
uency range.
Why? you ask. Back to Physics 1. When the cone is moving pistonically it is
basically acting as a mass on a spring. The mass involved includes the mass
of the cone, and the compliance includes the suspension compliance of the
woofer itself and compliance of the air trapped within the enclosure. Physics
tells us that a given mass and compliance will resonate at a certain frequency
at a certain damping. Also contributing to the damping of the system is the
electrical damping from the voice coil/ magnetic field system. The amount of
damping from this ciruit is theoretically unlimited. Thus, for a given cone
mass and resonant frequency, and system damping is possible. In general, there
is no problem designing a second order system with most woofers on the market
with optimal damping (Q=0.5 or 0.707 depending).
Of course the theory falls apart if the resoting forces are not linear. Most
speaker suspensions are fairly linear. Compression and rarefaction of air at the
proportions happening in most enclosures is VERY linear. So the theory stands.
I haven't dealt with 4th order loading, but the phenomena are virtually
identical.
I also have not dealt with the linearity of the voicce-coil/magnet system.
Many of these are not very linear. But that is irrespective of cone mass.
In general, woofer size (cone mass) will not effect low frequency "spped".
Mithat
Mike Hejl
> mass and resonant frequency, and system damping is possible. In general, there
> is no problem designing a second order system with most woofers on the market
> with optimal damping (Q=0.5 or 0.707 depending).
Excuse me, but how is a Q=0.5 optimal? A Q of .5 will have a substantial
bass rolloff with the -3db point above the system resonance and, in general,
will provide "weak" bass. Remember, most people don't want a speaker with a
Q as low as even .707 as evidenced by the overwhelming number of production
speakers with Qs well over 1.5.
According to Thiel-Small, et. al. equations for frequency response, a Q
of .707 gives the "flattest" response. Further, if I remember correctly
(I may not), Thiel specifies 1.0 (or is it 1.1) as "optimal" because of the
slight bass peak (<2db). Didn't Thiel even name this alignment "sub-
chebeychev" or some such? As a further example, suppose Vab for a particular
woofer is 1 cu. ft. for a Qb of .707. To obtain a Qb of .5, you'd have to
almost triple the box volume. (Remember, Vab follows an exponential
relationship to Qb. This isn't evident from the original papers, but just plot
the relationship between Qb and Vab.)
> In general, woofer size (cone mass) will not effect low frequency "spped".
It seems that Bl (the magnetic flux density in the voice coil gap) has
considerably more to do with cone excursion than *simply* cone mass. For
example, put a week magnet on a large cone mass and your theory falls apart.
Also consider a relatively short voice coil with little or no overhang.
Further, you don't even consider Qtc, the woofer's mechanical Q, nor a host of
other parameters associated with the woofer, enclosure, and even the damping
factor of the amp which is not, in reality, infinite.
Its obviously been a while since I looked at this stuff. However, Thiel-Small
has been around a long time with many diffrent "interpretations" in print.
Perhaps the best I've seen was written by Speaker Lab in Seattle, Wa. around
1980. (Anyone know if they are still around?) Conversely, the worst I've seen
are those "how to" books sold in places like Rat Shack. I'll dig up some of
this stuff if anyone is interested.
Mike
Francis Vaughan
|> In article <1991Mar3.2...@src.honeywell.com>,
ko...@lennon.SRC.Honeywell.COM (Mithat F Konar) writes:
|> > mass and resonant frequency, and system damping is possible. In
general, there
|> > is no problem designing a second order system with most woofers on
the market
|> > with optimal damping (Q=0.5 or 0.707 depending).
You need to distinguish between the Qts of the driver and the final Q of the
speaker system. From memory Thiel and Small showed how to build an optimal
box for any driver (within limits), but then showed that the most optimal
optimal design was one that started with a Qts for the driver of about 0.42.
Sealed systems like woofers with Qs of about the 0.7 mark.
|> Excuse me, but how is a Q=0.5 optimal? A Q of .5 will have a substantial
|> bass rolloff with the -3db point above the system resonance and, in general,
|> will provide "weak" bass. Remember, most people don't want a speaker with a
|> Q as low as even .707 as evidenced by the overwhelming number of production
|> speakers with Qs well over 1.5.
|>
|> According to Thiel-Small, et. al. equations for frequency response, a Q
|> of .707 gives the "flattest" response. Further, if I remember correctly
|> (I may not), Thiel specifies 1.0 (or is it 1.1) as "optimal" because of the
|> slight bass peak (<2db). Didn't Thiel even name this alignment "sub-
|> chebeychev" or some such?
Streching back into the memory, I seem to remember that the alignment most
preferred by Small was the QB3 alignment (Quasi 3rd order Butterworth) which
I think ended up looking a lot like a Q of around 0.7. I really need to
look this up however.
A lot of people are less concerened with the ultimate bass response and more
with transient response and a lack of passband ripple. Both improve with
lower Q. A Chebychev aligment gets you the best extention of bass at the
expense of a steep roll off, noticable ripple in response and fairly lousy
transient response. Butterworth is often taken as the "nice" compromise,
losing a little bass of the bottom but with no passband ripple, and reasonable
transient response.
A Q higher than 1.0 gets you real bass boom, but no real extention in response,
lots of ripple (hence the boom) and really lousy transient response. When you
say "the overwhelming number of production speakers" I suspect that this is
a reflection of the mass market in sound. The boom box market. Who remembers
the glorious scene in "Ruthless People" with the salesman trying to sell the
BIG speaker?
|> Its obviously been a while since I looked at this stuff. However,
Thiel-Small
Me too unfortunatly.
|> has been around a long time with many diffrent "interpretations" in print.
|> Perhaps the best I've seen was written by Speaker Lab in Seattle, Wa. around
|> 1980. (Anyone know if they are still around?) Conversely, the worst
I've seen
|> are those "how to" books sold in places like Rat Shack. I'll dig up some of
|> this stuff if anyone is interested.
Yeah, I have seen some absolutely abortionate efforts of how-toos. A
friend of mine tried to use the method in "Great Sound Stereo Speaker Manual"
by David Weems and got the most wierd results I have seen. When we tried the
algorithms directly from Smalls work the results were dramaticly different
(like an enclosure twice the volume.)
Francis Vaughan
Shankar Bhattacharyya
Mithat Konar is obviously talking about closed boxes. He is quite clear
about that. If memory serves, his original post just said "second order."
For them that don't know this already, 2nd order woofer = closed box.
From much that others have said, I suspect that some of them are not
talking about closed boxes, and some are not sure that there are design
distinctions betwen closed and vented boxes. There has been a certain
amount of mixing of terms, design criteria, and so on.
It would help if people said up front just what kind of woofer they are
assuming.
I take the liberty of summarizing some of what Mithat said, without
specific attribution, and without the accompanying quotes from earlier
posts. Maybe it is useful to put all of this in one place.
Here are the criteria which people use for the system Q of a closed box
system.
Assume a CLOSED box.
Total driver Q = Qts.
Total system Q = Qtc, for driver mounted in the box.
For Qtc <= 0.5 there is no ringing, i.e. there are no irregularities in the
impulse response, no oscillatory behaviour.
Impulse response freaks regard this as the maximum allowable Qtc. It gives
the greatest bass extension possible without ringing. Hence one kind of
optimum.
For Qtc <= 0.707 there is no ripple in the amplitude response, and maximally
flat response is achieved.
Bass extension freaks regard Qtc = 0.707 as the optimum.
Incidentally, higher Qtc does not produce lower F3, nor does lower Qtc.
Qtc = 0.707 gives the lowest F3 for any driver for which Qts < 0.707.
The related misconceptions come from the way in which Small's original
paper on closed boxes presents the dependence of F3 on box size, a
presentation which confuses many people. It took me a while to figure it
out.
If people are going to dispute this, I would urge them to make their
assumptions clear, and to remember that I am talking about closed boxes.
Let's not mix apples and oranges here.
For 0.5 < Qtc < 0.707 there is no ripple in the amplitude response, but
there is some oscillatory behaviour, i.e. some ringing.
For Qtc > 0.707 there is ripple in the amplitude response, and there is
ringing. The worst of both worlds, I suppose.
At Qtc = 1.3 the response ripple is about 3 db.
I lean towards Qtc = 0.707.
Also, someone had said something about box volume increasing exponentially
as some Q or other got smaller. What matters is not any particular Q, but
the term 1/(Qtc - Qts) ** 2, nitpicking issues aside.
If Qtc - Qts is a small positive number, box volume will be very large.
Qtc < Qts is not realizable.
People also make rather too much of a big deal over bass extension in
closed box woofers as related to the system Qtc. For a given driver, as
Qtc varies from 0.5 to over 1, the F3 only varies by about 1/5 octave.
That is not a whole lot. If you choose a decent driver of respectable size
and excursion, a little judicious equalization will take care of that.
Closed boxes are easy to design, and do not need religious approaches. If
I remember correctly, this is basically what Mithat said in his first post
on the subject.
Even quite high Qtc systems can be acceptable. Getting Qtc as low as 0.707
from a Dynaudio 30W100 takes 960 liters by my calculations. That is pretty
darned big. Yet it is used in the Duntech systems, I believe. Even though
the Duntechs are not small, I really doubt that they are large enough to
accommodate multiple 30W100's and still keep Qtc as low as 0.7 . The high
regard in which the Duntechs are held does suggest that getting obsessive
about Qtc may not be necessary.
If anyone tries hard enough, it remains possible to make truly bad woofers.
But decent closed box woofers are not hard to put together.
Ok, I've blown some steam. I'll go away now.
- Shankar
Bryan Ebersole
Vas (measured in cubic feet, in Parts Express catalog along with
all sorts of Qxx parameters)?
-----------
Other questions I have (my roommate and I are going to
try to subwoofer this summer so...)
What books do you recommend for building speakers/subwoofers?
(No, I don't want to read that Thiel paper--need a pratical book)
Is the book by Weems (the only one I have been able to find) ok?
What subwoofer drivers are recommended (brands)--Polydax has
an 11" woofer that goes down to 16 Hz (resonates at 17)--is this
for real, or is it one of those +/- 5000000 db type measurements
(again listed in Parts Express catalog, had a Vas ~ 27 ft^3)
Thanks for your input.
Bryan Ebersole
ig...@engin.umich.edu
(If this post blows up like my last one, U of Mich will have
one less news posting machine... :> )
Jonathan Fischer
>In article <1991Mar01.1...@sco.COM> jfis...@sco.COM (Jonathan Fischer) writes:
>> (1) Smaller cones have better transient response (speed).
>Er..no. Remember Physics 1 at college. You'll need to keep it in mind for
>the following discussion.
>
>being "slower" at low frequencies.]
Hmmmmm. I'm having a hard time buying this argument.
Say a speaker is reproducing a WHACK on a drum. The initial
attack will contain several components across the frequency spectrum;
the tweeter and the woofer/midrange will be responding simultaneously.
To keep the question intentionally simple: won't the lighter mass of
an 8" driver allow it to keep more in step with the tweeter than a
heavier 12 or 15"? This is what is meant by "fast".
Regardless of damping factor and resonances, a heavier object
will take longer to accelerate from rest to point "x" than a lighter
object, given the same force.
Mithat F Konar
>In article <1991Mar3.2...@src.honeywell.com>, ko...@lennon.SRC.Honeywell.COM (Mithat F Konar) writes:
>> mass and resonant frequency, and system damping is possible. In general, there
>> is no problem designing a second order system with most woofers on the market
>> with optimal damping (Q=0.5 or 0.707 depending).
>
>Excuse me, but how is a Q=0.5 optimal? A Q of .5 will have a substantial
>bass rolloff with the -3db point above the system resonance and, in general,
>will provide "weak" bass. Remember, most people don't want a speaker with a
>Q as low as even .707 as evidenced by the overwhelming number of production
>speakers with Qs well over 1.5.
A Qts of .707 will provide the lowest cutoff point for a given woofer with no
peaking in the amplitude response. The will be some overshoot in the time
response, though. Many feel that a second order system with a Q of 0.5 (which
will be six dB down at resonance rather than three for a Q of .707) is an
optimum balance of transient response and low frequency extension. Exactly
how that number was arrived at I can't remember. It is afairly common one,
however.
A system Q of 1.5 will have considerable peaking at resonance and considerable
overshoot and ringing. This gives the impression to many of lots of bass. On
further listening, however, and by camparison to an optimal alignment, one
discovers that this "abundant" bass is not very tight and not very distinct.
One aspect of second order systems that I didn't explicate as well as I might
have is that, like a mass on a spring that doessn't oscillate forever, speakers
have damping factors. The damping comes from two sources: the mechanics of
the cone and suspension and the electromechanical damping from the voicecoil/
magnet assembly. Typically, the mechanical damping is an order of magnitude
less than the electromagnetic (i.e: Qts is about ten times Qes). Thus,
the majority of a speaker's damping comes from the voicecoil/magnet assembly.
If you assume a speaker having a voicecoil d.c. resistance of 6ohms, and
an amplifier with an output impedance of .3ohm (not at all uncommon for
solid state amps), the change in damping as a result of the amplifier's
output impedance is almost negligible. Tube amps, however, can have quite
high output impedances. This may account for some of the "warmth" tube-freaks
love in tube amps.
Mithat Konar
Mithat F Konar
> Say a speaker is reproducing a WHACK on a drum. The initial
>attack will contain several components across the frequency spectrum;
>the tweeter and the woofer/midrange will be responding simultaneously.
>To keep the question intentionally simple: won't the lighter mass of
>an 8" driver allow it to keep more in step with the tweeter than a
>heavier 12 or 15"? This is what is meant by "fast".
>
> Regardless of damping factor and resonances, a heavier object
>will take longer to accelerate from rest to point "x" than a lighter
>object, given the same force.
Given the same force, a lighter object will respond "more quickly." Who said
the force was the same, however?
Low frequency behavior of second order systems cam be described completely
with two parameters: resonant frequency and damping. If you take a system
with a given cone mass, compliance, etc. and increase the cone mass, the
resonant frequency will drop and the system damping will be reduced. To
restore the damping to the previous amount, you can increase the force
factor (Bl product) coming from the magnet system.
One must keep clear whether one is speaking of HIGH FREQUENCY PHENOMENA or
LOW FREQUENCY PHENOMENA, though. Generally speaking, a higher mass woofer will
have a lower high frequency cutoff. What does that mean? It means that it
has a smaller useable bandwidth and must be crossed over sooner to a midrange
or tweeter. Does that mean that the woofer won't "keep up" with the tweeter?
The short answer is no. This "mass limited" (which is a horrendous over-
generalization to begin with, but..) high frequency rolloff is just that: a
high frequency rolloff that can be modelled mathematically using transfer
functions or electrically using capacitors, resistors and inductors. This
high frequency roloff can then be incorporated into a filter design in
conjunction with a crossover network (i.e. the system response is a combination
of the drivers' responses and the characteristics of the networks feeding
them).
The bottom line is that IN THE HIGH FREQUENCIES if a speaker doesn't respond
quickly because of a mass related issue, this manifests itself in a plain
old simple high frequency roloff. It doesn't respond "quickly", nor does it
need to. If you look at the edges of a pulse that's been subject to HF
rolloff, you will see more-or-less a pulse with sloped and rounded edges
rather than vertical and perfectly square. That's what comes out of the woofer.
It's the tweeter's job to fill in the lost harmonics so the edges return to
vertical and square. Most crossovers will muck up the phase resonse so that
by the time you're done, the pulse will look nothing like a pulse. But
you will have that problem as long as you have a crossover.
If this still seems a bit puzzling, perhaps a brief look into Fourrier analysis
will help.
Mithat Konar
david.r.moran
but be advised that once you put the driver and its enclosure into
an enclosed space to strike the air therein (stick it in a room,
in other words), virtually all of the subtleties of
response and design are swamped by boundary augmentation, ripple/
lack of ripple due to good/bad radiation loading, room resonances due to
proportions, etc. etc. etc.
If you really want to do the design yourself, though, the most elegant
and user-friendly software I have heard of is from SpeakEasy in Newton,
Mass., called LF Designer
or some such name, and is advertised in the current AESJ. It is not altogether
a unique program but it is thorough and covers all of the variables
exhaustively, with popup subscreens, helptext, and the like. Under $200,
as I recall. The ad is worth checking out for its graphics alone.
Francis Vaughan
(Shankar Bhattacharyya) writes:
|> Time for a meta-post.
[Deleted a very well written and useful summary of the state of the art]
|> Even quite high Qtc systems can be acceptable. Getting Qtc as low as 0.707
|> from a Dynaudio 30W100 takes 960 liters by my calculations. That is pretty
|> darned big. Yet it is used in the Duntech systems, I believe. Even though
|> the Duntechs are not small, I really doubt that they are large enough to
|> accommodate multiple 30W100's and still keep Qtc as low as 0.7 . The high
|> regard in which the Duntechs are held does suggest that getting obsessive
|> about Qtc may not be necessary.
I once worked out the volume of the Soveriegns enclosure and it comes up pretty
close to 900 litres per driver, but still short of 960 from memory.
John Dunlevy claims that a Qtc of 0.82 is optimal, I don't know why as he
used to design to 0.707. The Sovereigns really are a big speaker, they are
much deeper that you might think.
|> If anyone tries hard enough, it remains possible to make truly bad woofers.
|> But decent closed box woofers are not hard to put together.
Very true. It is probably one of the best understood aspects of speaker design.
Francis Vaughan
Francis Vaughan
(Bryan Ebersole) writes:
|> What do all these Qxx parameters mean in real life, and what is
|> Vas (measured in cubic feet, in Parts Express catalog along with
|> all sorts of Qxx parameters)?
|> -----------
Sometime ago Shankar Bhattacharyya ( sbha...@rnd.gba.nyu.edu )
posted a few messages that gave a quick overview. I saved these
but perhaps, if he is still reading, and wishes too, he could repost
a pertinent summary.
|> Other questions I have (my roommate and I are going to
|> try to subwoofer this summer so...)
|> What books do you recommend for building speakers/subwoofers?
|> (No, I don't want to read that Thiel paper--need a pratical book)
|> Is the book by Weems (the only one I have been able to find) ok?
No. The speaker design stuff in here supposedly based on Theils and
Smalls work seems to be broken. I don't know what is wrong but the results
you get out are weird.
There are a few good papers published by Small. These are essentially
trying to demystify Theil.
Closed-Box Loudspeaker Systems, Part I: Analysis,
Journal of the Audio Engineering Society, vol 20, # 10, p 798 (1972)
Closed-Box Loudspeaker Systems, Part II: Synthesis,
ibid, vol 21, # 1, p 11 (1973)
The synthesis paper is good, but you need the analysis paper to
understand the terms. It still needs some work to understand the papers.
|> What subwoofer drivers are recommended (brands)--Polydax has
|> an 11" woofer that goes down to 16 Hz (resonates at 17)--is this
|> for real, or is it one of those +/- 5000000 db type measurements
|> (again listed in Parts Express catalog, had a Vas ~ 27 ft^3)
Its going to be a big box.
A driver like this is quite possible. It may not be real efficient.
I used to own a couple of brilliant drivers, the IMF bass drivers
they used in their mk IV transmission lines. Perfect for a ported
enclosure, flat to 20hz. (Not kidding.) I had to sell them. Pity.
Francis Vaughan.
Shankar Bhattacharyya
>Sometime ago Shankar Bhattacharyya ( sbha...@rnd.gba.nyu.edu )
>posted a few messages that gave a quick overview. I saved these
>but perhaps, if he is still reading, and wishes too, he could repost
>a pertinent summary.
OK, I will. I don't have the time to clean it up and make a decent summary,
so I will just do a casual job of picking stuff from the articles. I hope
that serves.
> Closed-Box Loudspeaker Systems, Part I: Analysis,
> Journal of the Audio Engineering Society, vol 20, # 10, p 798 (1972)
>
> Closed-Box Loudspeaker Systems, Part II: Synthesis,
> ibid, vol 21, # 1, p 11 (1973)
>
>The synthesis paper is good, but you need the analysis paper to
>understand the terms. It still needs some work to understand the papers.
I think the original "Direct-Radiator Loudspeaker Systems" paper is also
useful towards getting to know the terminology, but obviously less
important.
Till recently, there was nothing to suggest that anyone wanted to do
anything practical. Since some people seem interested in practical
application, here goes.
Late last summer I posted a series of articles on us.high-end, on closed
box woofers. Those are archived by Tom Krueger. Since they are available
by ftp from his machine, I won't post all that stuff again. I will just
post formulae, and results from applying them to a few drivers.
If you want more detail, please look into the archives. If your site does
not have the technology for ftp to internet sites, I can mail the files
to you, but I don't plan to repost it all again. It would seem like a
bit of a waste.
In any case, the articles were not aimed at experienced builders, and any
decent book has the basic information.
I had one article apiece on: feasibility and realistic expectations;
references; formulae; good drivers; and drivers which look good initially,
but turn out to be less satisfactory than one might expect.
Seth Bradley and Bill Spencer posted designs, and there was a fair
amount of traffic on the subject. Seth's design was a sixth order vented
system, with admirably high output. Bill's was a sealed box. You may
find the archives useful. Bill also posted a list of places to get parts
and supplies, and has posted similar information here.
Part 3 of that set of articles had the formulae, including a typo. In one
place, it read "*8" where it should have read "**". Once you are aware
of this, it is glaringly obvious. I missed it, obviously. Just a
warning.
I'm a bit busy just now, so my ability to follow up on this is limited.
Sorry about that. I'll do what I can. Besides, I'm just a semi-informed
builder, not an expert. However, many people on the net know much more
about this than I do, so there should be no shortage of information.
Just by the way, Zalytron is selling the Dynaudio 30W54 for $99 right now.
Good price, but the Audio Concepts AC-12 remains very attractive, at lower
cost, even if it has less sex appeal.
As an aside, I favour Qtc of about 0.7. For that you want a Fs of about
20 hz or so, and Qts about 0.45. Don't get all nitpicky about this
statement; it's a throwaway.
----------------------------------------------------------------------
Shankar Bhattacharyya, Information Systems, New York University
sbha...@rnd.gba.nyu.edu
----------------------------------------------------------------------
Shankar Bhattacharyya
several drivers. Experienced builders already know everything in this article.
These results are for CLOSED boxes only. They assume radiation into 4*pi,
and assume small signals. As David Moran has pointed out, this has little
to do with how systems behave in rooms.
If you find any errors, please let me know.
This is long, more than 350 lines, and of no use to anyone who is not
interested in building woofers, and probably not of any interest either.
NOTATION:
Fs = driver free air resonance frequency
Qes = driver Q at Fs, counting electrical resistance only, free air
Qms = driver Q at Fs, counting non-electrical resistance only, free air
Qts = total driver Q at Fs, free air
(let's not pick nits about the "at Fs")
EBP = efficiency bandwidth product = Fs / Qes
Vas = compliance of unbaffled driver, expressed as volume of air with
equal compliance
Sd = effective piston area of driver
Xmax = maximum linear excursion
Vd = maximum linear displacement = Sd * Xmax (Small's usage)
parameters for a closed box system:
Vb = net internal volume of enclosure
Fc = system resonance frequency, with driver mounted in the enclosure
F3 = half power frequency, cutoff frequency, -3 db point
Qtc = Q of complete loudspeaker system, i.e. of driver "plus" enclosure
alpha = compliance ratio = Vas / Vb (approximation)
The approximation holds reasonably closely if the box is unfilled.
I'm going to assume that it is good. Stuffing the box makes the enclosure
seem larger, and you would have to determine what the effective internal
volume is (compliance, actually).
Cautions:
Xmax is intended to be given as an amplitude, i.e. one way. Dynaudio
expresses it as peak-to-peak, or twice the maximum amplitude.
Small seems to want Vd expressed as a one way number. Some manufacturers
seem to use it for the peak-to-peak number.
It is entirely possible that some manufacturers may specify Xmax as the
one-way displacement, but Vd as a peak-to-peak number. I don't know.
So read specifications with caution.
SPL CALCULATIONS:
This information is straight out of one of Linkwitz's articles. I presume
it is correct. I include this just in case someone is interested. If you
work through it, it is instructive to see why woofers are expensive. It
assumes a pulsating sphere, i.e. a speaker small compared to the
wavelength, and radiation into 4*pi stearadians.
If you get interested in power handling, and output spl, I suggest that you
look at the literature. I don't know any more than this on the subject.
SPL in dB = 94.3 + 20logx + 40logf + 40logd - 20logr
where x = amplitude of the oscillation, (one way)
d = effective diameter of woofer (actual radiating diameter)
r = distance between loudspeaker and measuring point
f = frequency
frequency in hertz, x, d, r in meters.
CLOSED BOX FORMULAE:
EBP = Fs / Qes
Small says that drivers for which EBP is about 50 are good for closed boxes,
and about 100 for vented boxes.
If you don't have Qes, assume it is about 1.1 X Qts.
alpha = Vas / Vb
= (Fc / Fs) ** 2 - 1
= (Qtc / Qts) ** 2 - 1
term = - 2 + 1/(Qtc ** 2) (just for simplicity)
( My original post on us.high-end had a typo on that line. It had
"*8" where the "**" should be.
So, if you ftp the original articles, do make that correction. )
F3 = Fc * sqrt { 0.5 * [ term + sqrt(term * term + 4) ] }
If you shoot for Qtc only slightly higher than Qts, you will need a huge
box. For many drivers you will find that the required volume declines very,
very fast as you shoot for slightly higher Qtc. Qtc = 0.8 has about 0.25 dB
ripple, 0.9 has about 0.5 dB, 1.0 has about 1 dB. I think that 0.5 dB
ripple is quite tolerable, so, if dramatic size reductions can be realized
by accepting slightly higher Qtc, I would accept it happily.
------------------------------------------------------------------
Here are numbers for closed box alignments for a few drivers which look like
good candidates for home made woofers.
The numbers were generated by a C program of about 20 lines of code, which
I can no longer find, but which should be easy enough to write from
scratch. It took me half an hour the last time. I presume that anyone on
the network can do this too. I'm no hot-shot programmer.
I suggest that you check numbers out before you believe them completely.
I looked at every woofer in the Madisound and Audio Concepts catalogues,
choosing those which loked like they had acceptable EBP (not too far from
EBP = 50). PRICES MAY BE OUT OF DATE. This file is quite old.
I tried to stick to drivers with polymer cones (much more rugged, and
possibly better able to work as a rigid piston). With one or two
exceptions, I have tried to avoid vastly expensive drivers. Dynaudio
and Eton drivers are very highly regarded, so I have included some of
those, including a couple which cost over $200 just by themselves.
Finally, I have tried to restrict myself to drivers which are capable of
producing serious amounts of bass in a closed box design. So I have
stayed away from 6" drivers and other such things. Most of these drivers
are 10" or 12" units. There are a couple of smaller ones.
The Audio Concepts AC-10 and AC-12 look far better than anything else I
have seen from either the Audio Concepts or Madisound catalogues.
Some drivers which do not look good in closed boxes will work just fine in
vented ones. Eton drivers, for instance.
All frequencies are in hertz, volumes in liters.
Incidentally:
1 liter = 61 cu in
1 cu ft = 28.3 liters
Dynaudio 24W100 $ 110 looks good, but expensive, for a small system
Qts = 0.35, Vas = 132.00, Fs = 23.0, Qes = 0.45, EBP = 51.11
qtc alpha Vb Fc F3
(hz) (hz)
0.50 1.04 126.82 32.86 51.05
0.60 1.94 68.08 39.43 47.67
0.70 3.00 44.00 46.00 46.47
0.80 4.22 31.25 52.57 47.16
0.90 5.61 23.52 59.14 49.06
1.00 7.16 18.43 65.71 51.66
1.10 8.88 14.87 72.29 54.70
1.20 10.76 12.27 78.86 58.02
1.30 12.80 10.32 85.43 61.54
Dynaudio 30W54 $ 150 works very well with some bass boost.
Xmax is less than I would like, but may be
Expensive, but conservatively stated. My 30W54 based
possible within woofer makes quite enough noise for me.
a budget of $200 Rolls Royce grade construction, excellent
rise time.
Qts = 0.36, Vas = 254.00, Fs = 22.0, Qes = 0.42, EBP = 52.38
qtc alpha Vb Fc F3
(hz) (hz)
0.50 0.96 264.15 30.81 47.88
0.60 1.82 139.20 36.97 44.71
0.70 2.84 89.29 43.14 43.58
0.80 4.02 63.16 49.30 44.23
0.90 5.36 47.43 55.46 46.00
1.00 6.85 37.10 61.62 48.45
1.10 8.49 29.90 67.79 51.30
1.20 10.30 24.66 73.95 54.41
1.30 12.26 20.72 80.11 57.71
Dynaudio 30W100 $ 240 looks quite promising; if you can accept a
Qtc of 1.2, with its attendant 2 db respone hump
Included because in the midbass, this is a hell of a driver.
people are inte- Rolls Royce grade construction, huge voice coil,
rested in Dynaudio excellent rise time. Used in the Duntechs, I
Too expensive for believe.
a budget woofer.
Qts = 0.62, Vas = 269.00, Fs = 24.0, Qes = 0.80, EBP = 30.00
qtc alpha Vb Fc F3
(hz) (hz)
0.50 requires non-positive box volume -- not feasible
0.60 requires non-positive box volume -- not feasible
0.70 0.28 964.73 27.14 27.42
0.80 0.67 401.30 31.02 27.83
0.90 1.11 241.47 34.89 28.94
1.00 1.61 167.09 38.77 30.48
1.10 2.16 124.66 42.65 32.27
1.20 2.76 97.53 46.53 34.24
1.30 3.41 78.87 50.40 36.31
Audio Concepts AC-10 $ 50 Looks very good. One of only a few drivers
of reasonable cost which will produce
Xmax = +/- 7mm F3 below 40 hz in a closed box without
bass boost. Excellent excursion. Among
10" woofers, this would probably be my
choice.
Qts = 0.44, Vas = 138.00, Fs = 23.0, Qes = 0.50, EBP = 46.00
qtc alpha Vb Fc F3
(hz) (hz)
0.50 0.29 473.70 26.14 40.61
0.60 0.86 160.56 31.36 37.92
0.70 1.53 90.14 36.59 36.97
0.80 2.31 59.85 41.82 37.52
0.90 3.18 43.34 47.05 39.02
1.00 4.17 33.13 52.27 41.09
1.10 5.25 26.29 57.50 43.51
1.20 6.44 21.44 62.73 46.16
1.30 7.73 17.85 67.95 48.95
Audio Concepts AC-12 $ 60 This looks stunningly promising. If I
were building a woofer today, this is what
Xmax = +/- 8 mm I would use. 30 hz, at respectable
excursion, without bass boost. The numbers
look much better than for the Dynaudios.
Len Moskowitz, with real software, gets
numbers which are not quite as good. But
this remains most promising. Cheap, too.
Qts = 0.43, Vas = 240.00, Fs = 18.0, Qes = 0.49, EBP = 36.73
qtc alpha Vb Fc F3
(hz) (hz)
0.50 0.35 681.66 20.93 32.52
0.60 0.95 253.43 25.12 30.37
0.70 1.65 145.45 29.30 29.60
0.80 2.46 97.51 33.49 30.04
0.90 3.38 70.99 37.67 31.25
1.00 4.41 54.44 41.86 32.91
1.10 5.54 43.29 46.05 34.85
1.20 6.79 35.36 50.23 36.96
1.30 8.14 29.48 54.42 39.20
Madisound 1252DVC (dual voice coil) $ 40
Looks quite promising, but I have no idea how I should deal
with the dual voice coil design. My numbers for this may be
entirely wrong. But not as promising as the AC-12, or even the
AC-10. Note the size required for Qtc=0.7. I would shoot for
Qtc=0.8 with this driver.
The Qts looks more suited for a vented box, but I think the
box would be quite large. If you can live with that, this
would produce very deep bass (guestimate; I haven't run the
numbers.
Qts = 0.36, Vas = 318.00, Fs = 19.0, Qes = 0.39, EBP = 48.72
qtc alpha Vb Fc F3
(hz) (hz)
0.50 0.93 342.30 26.39 41.00
0.60 1.78 178.88 31.67 38.29
0.70 2.78 114.35 36.94 37.32
0.80 3.94 80.75 42.22 37.88
0.90 5.25 60.57 47.50 39.40
1.00 6.72 47.35 52.78 41.49
1.10 8.34 38.15 58.06 43.93
1.20 10.11 31.45 63.33 46.60
1.30 12.04 26.41 68.61 49.42
Eclipse W1238R $ 50 Not bad, but I would prefer the AC-12.
As with the Madisound driver, the Qts
suggests a vented box, although the EBP, as
with the Madisound, looks rather low for
a vented box.
Qts = 0.33, Vas = 280.00, Fs = 19.0, Qes = 0.37, EBP = 51.35
qtc alpha Vb Fc F3
(hz) (hz)
0.50 1.30 216.10 28.79 44.73
0.60 2.31 121.43 34.55 41.77
0.70 3.50 80.01 40.30 40.72
0.80 4.88 57.41 46.06 41.32
0.90 6.44 43.49 51.82 42.98
1.00 8.18 34.22 57.58 45.26
1.10 10.11 27.69 63.33 47.93
1.20 12.22 22.91 69.09 50.84
1.30 14.52 19.29 74.85 53.92
The following drivers look good initially, or are highly regarded in
various ways. they do not work well in sealed boxes. Many will work very
well in vented boxes, although I have not attempted any design work.
Dynaudio 24W75 $ 50
Qts = 0.83, Vas = 88.00, Fs = 33.0, Qes = 1.11, EBP = 29.73
Qtc < 1.2 is not convenient. I don't see why this woofer was ever made.
I don't see how it can work in any kind of box.
JC12 (from Audio Concepts) paper cone, dual voice coil $ 60
Qts = 0.31, Vas = 339.00, Fs = 18.0, Qes = 0.32, EBP = 56.25
You cannot get F3 = 40 hz in a closed box. However, I don't know how
to deal with the dual coils, so my conclusions may be wrong. Looks good
for a vented box.
Audio Concepts AC-8 $ 40
Qts = 0.41, Vas = 58.00, Fs = 30.0, Qes = 0.50, EBP = 60.00
Vb = 31 liters gives Qtc = 0.7, F3 = 52 hz.
Actually, this looks quite good in a closed box. The issue is that since we
are talking about woofers to extend bass into serious territiory, 50+ hz
just does not do it. But this would be a serious candidate for a small two
way system.
Swan 305 (substitute for Precision/Peerless TA305) $ 50
Qts = 0.31, Vas = 167.00, Fs = 24.5, Qes = 0.34, EBP = 71.22
This looks terrible in a sealed box. In a vented box, however, it should
perform very well, with F3 in the 20's. I included this basically because
people will think of the TA305. Note that the EBP is high for a closed box.
Seth Bradley's design uses the original TA305 in a vented box, very
successfully. In a sealed box you cannot get F3 < 55 hz. Seth's design will
allow you to break your lease with no effort at all.
KEF B139 $ 108
Qts = 0.37, Vas = 164.00, Fs = 25.0, Qes = 0.40, EBP = 62.50
This driver is something of a legend, particularly in Britophile circles.
I recall seeing it in an early WAMM, and it has been used in many
loudspeakers, including, obviously, many from KEF. However, if I were
spending $100 on a driver, I would be annoyed if I could get no better than
48 hz bass without equalization. I think this would work much better in a
vented box, yielding F3 in the mid 20's perhaps. It is also a popular
driver in the builder community for transmission lines. Until the Dynaudios
became well-known in the US, this used to be the fancy driver of choice.
Many people still use it.
Eton 11-580 $ 210
Qts = 0.30, Vas = 180.00, Fs = 23.0, Qes = 0.31, EBP = 74.19
I include this because Etons are well regarded, and people will consider
them. This is an 11" driver. For $210 you get F3 > 50 hz. This is obviously
silly. Note the low Q, and consequent high EBP. This was designed for a
vented box, and should work very well in one.
----------------------------------------------------------------------
Shankar Bhattacharyya, Information Systems, New York University
sbha...@rnd.gba.nyu.edu, sbha...@sales.gba.nyu.edu
----------------------------------------------------------------------
Biswa Ranjan Ghosh
"fast" versus "slow" woofers. I just want to add my two cents worth in
the same vein...
If one makes the underlying assumption that drivers operate in a linear and
time-invariant fashion, that is, the output SPL is a linear function of the
input voltage, and doesn't depend on the time of day, then it is a "linear
system," and one can make use of results from linear systems theory. One
of these is that the driver is characterized by its output spl vs. input
voltage frequency response, measuring both magnitude and phase. Other
measurements, such as step and impulse response, are *completely* tied to
the frequency response, and any one of them can be derived from any of the
others. This means, among other things, that the transient response is
completely predicted from the frequency response. It also means that the
low-end characteristics, like damping and rolloff, can be seen from the
frequency response when the driver is mounted in a specific enclosure.
All this talk about "fast" vs. "slow" woofers is just technobabble. You
say your woofer is "faster" than mine? I say, "Big deal, all you're really
saying is that your driver's high-end cutoff frequency is higher than mine."
Not that it isn't a desirable characteristic, it certainly is. But it's
nothing new. If on the other hand, by "faster" you mean something about the
low end rolloff, then I say, "Well, why don't you talk about system Q and
low-end cutoff frequncy, which are the standard ways of characterizing the
low end response?" Given any woofer (within reason), you can easily find
a closed-box volume which will give you Q's of 0.5 to 1.0 (the spectrum
from tight, controlled rolloff to fat bass). Of course, each will have a
different cutoff frequency.
Next time a hi-fi salesman talks about slow woofers not keeping up with
the tweeters, find another place to shop! Or at the very least, demand
that he define the terms "fast" and "slow" in terms which actually mean
something.
Mike Hejl
> In article <6...@necssd.NEC.COM>, he...@necssd.NEC.COM (Mike Hejl) writes:
> |> In article <1991Mar3.2...@src.honeywell.com>,
> ko...@lennon.SRC.Honeywell.COM (Mithat F Konar) writes:
> |> > mass and resonant frequency, and system damping is possible. In
> general, there
> |> > is no problem designing a second order system with most woofers on
> the market
> |> > with optimal damping (Q=0.5 or 0.707 depending).
>
> You need to distinguish between the Qts of the driver and the final Q of the
> speaker system. From memory Thiel and Small showed how to build an optimal
Agreed. I assumed, possibly incorrectly, that Mithat was talking about box
Q (Qb). Besides, its unclear from your attribution who you are addressing
here. Its not me.
> box for any driver (within limits), but then showed that the most optimal
> optimal design was one that started with a Qts for the driver of about 0.42.
Thats about right. Qts=.39 is optimal for a *4th* order alignment and the
optimal Vb will equal Vab. This also gives minimum ripple.
> Sealed systems like woofers with Qs of about the 0.7 mark.
I'm not shure about this. Do you mean Qts? If so, if you have Qts = Qb, the
formulas will give you absurdly high Vab. Further, if Qts > Qb, the formula
will give you a NEGATIVE Vab. Just about ANY woofer can be used in a 2nd-order
alignment as long as Qts < Qb. It seems you mean Qb since Qb=.707 gives
optimal frequency response.
Perhaps we should concentrate on particular ALIGNMENTS. Things are getting
rather confusing with both sealed and vented alignments discussed in the same
post.
> |> Its obviously been a while since I looked at this stuff. However,
> Thiel-Small
>
> Me too unfortunatly.
All this has given me the itch to go home and do some research on this stuff
and start doing some building again. Unfortunately, I no longer have access
to a table saw and a circular saw will have to suffice. Maybe I'll have to
try a transmission line design again since Sony has "rediscovered" its
virtues. There is a local guru here who has been building *GREAT* TL and
dipole speakers for 40 years!
jeff waldeck
I have actually experienced this effect!
When in college, I built a set of subs using 12" woofers in 6 cu. ft. cabinets.
Using a sound meter, The loudest I could drive a 40 hz test tone was 108 db without causing adverse sounds (overdriving voive coil).
Then one day I read this and tried it out. I turned the subs around to face the
wall (appx 4" away). Using the same source, I could reach 112 db before experiencing the same sound (noise)
Of course, this was not a SCIENTIFIC test, but music could be played MUCH
louder with less distortion and less power.
jeffw@hpdmmfo